From the State Government edition of September 24, 2026
Digital States 2026 finds personal-productivity AI implemented in 76% of states as adoption becomes infrastructure
Center for Digital Government / Government Technology · State government IT / multi-state CIO survey · United States (50-state Digital States Survey)
- Publisher
- Government Technology
- Original publication
- September 24, 2026
- Source retrieved
- 2026-09-25
- Event date
- 2026-09-24
What happened
Why now: published today (2026-09-24), the biennial Digital States Survey documents a sharp shift from 2024 evaluation-heavy GenAI postures to 2026 production use. Personal-productivity AI implementation rose from 21% of states in 2024 to 76% in 2026 (remaining 24% on a path to enterprisewide tools). AI-augmented application development moved to 53% implemented (plus 38% pilots), while AI-assisted contract review rose from 2% to 22% implemented with 49% in pilots. Cyber budget share of 6–10% of IT spend rose from 28% (2024) to 43% (2026). The survey also surfaces state-level operating examples (Utah Gemini scale, Michigan benefits document automation, Iowa procurement chatbot) while noting persistent workforce gaps (84% need cyber staff; 76% need AI modeling staff). This is multi-state infrastructure evidence, not a single-agency ROI study.
Why it matters
Gives state CIOs and SLED sellers a same-day benchmark for how far enterprise GenAI, contract-review AI, and cyber/AI staffing gaps have moved across peer states—useful context beside single-agency chatbot metrics.
Evidence and measured results
GovTech/Center for Digital Government Digital States 2026 feature dated September 24, 2026, summarizing survey percentages and naming A-grade states with illustrative deployments. Figures are survey-reported, not independently audited outcome evaluations.
Limitations and uncertainty
Digital States is a vendor-adjacent Center for Digital Government survey published by GovTech (e.Republic). Percentages are self-reported by states; named agency outcomes (hours saved, automation rates) are illustrative and should be verified against primary agency sources before contracting claims.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-09-24; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- State IT leaders need peer benchmarks to justify enterprise GenAI scale vs continued pilots amid cyber and staffing constraints.
- Stakeholders
- State CIO / CDO
- AI CoE lead
- CISO
- Budget/legislative staff
- Agency digital service leads
- Discovery questions
- Which GenAI categories are implemented vs piloting here?
- What metrics does the legislature already accept?
- Where are approved-tool gaps creating shadow AI?
- How are cyber and AI staffing gaps prioritized?
- Value hypothesis
- A short benchmark-and-gap engagement can translate Digital States peer percentages into a prioritized roadmap without overclaiming survey ROI.
- Potential engagement
- 3-week Digital States gap assessment vs the state’s current GenAI inventory and metrics.
- Unsupported claims
- 76% implementation equals proven ROI
- A-grade status predicts vendor fit
- Survey anecdotes transfer as this state’s metrics
Pre-sales engineering
Role takeaway
- Fit
- Fits shared enterprise GenAI access, coding assistants, and contract-review services behind existing identity and SOC tooling.
- Architecture and integration
- Integrate approved copilots and analysis services with enterprise IdP, DLP, and content grounding; separate productivity from high-risk decisional AI.
- Prerequisites
- Approved GenAI policy and tool list
- Enterprise identity/SSO
- Logging/retention for AI prompts where required
- Data classification baseline
- Constraints
- Survey is multi-state and not jurisdiction-specific
- Federal policy churn may reorder priorities
- Workforce gaps may limit sustained ops
- Security
- Require approved-tool only access, SOC visibility for AI-assisted cyber workflows, and no training on non-public data without contract controls.
- Proof of value
- Inventory coverage of employees on approved tools
- Pilot-to-production conversion rate
- Contract-review cycle-time sample with human review retained
Delivery
Role takeaway
- Work
- Facilitate inventory, prioritize 1–2 production GenAI use cases, and wire governance checkpoints used by peer A states.
- Dependencies
- CIO sponsorship
- CISO approved-tool path
- Agency product owners
- Training/change capacity
- Ownership
- State CIO / AI CoE owns platform standards; agencies own use-case outcomes.
- Skills and adoption
- Pair tool access with role-based prompt/use training and CoE office hours; track weekly active use, not license counts alone.
- Governance checkpoints
- Use-case risk tiering
- Approved-tool exception review
- Public-facing disclosure check
- Cyber logging verification
- Acceptance criteria
- Documented GenAI inventory
- At least one production use case with defined metric
- Shadow-AI reduction plan
- Risks
- Survey-driven overbuying
- Metric cherry-picking from peer anecdotes
- Cyber/AI staff shortages delaying sustainment
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
States treating GenAI as infrastructure need shared platforms for productivity tools, app-dev assistants, and contract-review workflows, plus data-classification layers before agentic/data-management AI. Expect hybrid commercial LLM suites behind agency tenancy controls rather than one-off pilots.
Governance
Who approves, reviews and stays accountable for outcomes?
Survey framing ties AI rollout to cybersecurity, federal policy adaptability, and data governance maturity. CIOs should inventory approved tools and require funding/outcome reporting paths (as Georgia and Ohio examples illustrate) when moving from labs to production.
Security and privacy
What data, permissions and controls need testing?
Rising cyber budget share and AI-in-SOC narratives imply GenAI deployments must sit inside enterprise identity, logging, and approved-tool policies; shadow productivity AI remains a residual risk where 24% of states have not finished enterprisewide rollout.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
76% personal-productivity implementation implies large training and digital-equity workloads for employees with uneven AI literacy; 84%/76% cyber and AI modeling staffing needs signal hiring/upskilling pressure.
Procurement
What should contracts, pricing and exit terms secure?
Contract-review AI moving to 22% implemented / 49% piloting creates demand for evaluable vendor claims, SOC2/bias evidence, and state-approved model lists rather than one-off SaaS buys.
Operating model
Which teams own the service once it runs?
Survey theme is GenAI intertwined with cyber, modernization, and data governance—favor CoE + agency product ownership over isolated IT experiments.
Publication history
- 2026-09-24State Government · Issue 194 resources
Stable resource ID: digital-states-2026-ai-experiment-to-infrastructure-20260924